Qualitative Assessment of Local Tongue Deformation Using Principal Component Analysis in Patients Who Underwent Partial Glossectomy

Abstract

Purpose: The purpose of this study was to investigate patterns of muscle shortening in the sample of patients who underwent partial glossectomy and within a sample of healthy control subject using Magnetic Resonance Imaging (MRI) techniques and analyzing the data using the principal component analysis (PCA). Materials and Methods: PCA was used to analyze muscle shortening data collected via MRI imaging from eight control and five glossectomy individuals. The muscle length measurements were performed on the following five tongue muscle segments: 1) genioglossus posterior (GGp), 2) geniohyoid (GH), 3) transverse anterior (Ta), 4) transverse middle (Tm), and 5) transverse posterior (Tp). Results: Results: In the anteroposterior (x) direction, GGp and GH underwent the greatest amount of shortening for all three sounds. However, GH underwent greater shortening for /s/ while GGp underwent the greatest shortening for /u/ and /k/. In the mediolateral direction, Ta underwent the greatest shortening into /s/ and Tm and Tp underwent the greatest shortening into /u/ and /k/. Controls had less variance in shortening patterns than the patients. Conclusion: GH seems to be the primary protractor of the tongue for the production of the front sounds such as /s/. GGp seems to be the primary muscle that allows deformation of the base of the tongue, allowing it to be raised for production of velar sounds such as /uk/. In glossectomy patients, adjacent or contralateral muscles appear to be recruited as a compensatory mechanism for production of the same sounds as in controls. That being said, PCA was not able to differentiate between the controls and glossectomy patients which could be due to a very small sample size.